gmane.comp.editors.vim.vim-ruby.devel

Re: rubycomplete unusual gem behavior?

Mark Guzman — 2011-05-02T03:13:52

On Fri, 29 Apr 2011 12:08:23 -0700, Felix Gallo <felixgallo< at >gmail.com>
wrote:

I can't replicate this with a similar setup. Please confirm that vim is
actually using the ruby install
you expect. 

Here are some vim command outputs you should probably provide if that
isn't the issue:

version
ruby puts RUBY_VERSION
ruby puts $:
messages


  --mg

Re: rubycomplete unusual gem behavior?

Mark Guzman — 2011-04-30T19:52:00

On Fri, 29 Apr 2011 12:08:23 -0700, Felix Gallo <felixgallo< at >gmail.com>
wrote:

I'm guessing this was run from a script or irb, I'll take a look but
rubycomplete operates
inside Vim's embedded ruby environment which has some unique quirks last I
checked.


I should just toss in "hic sunt dracones" as a prefacing comment. 
  --mg

rubycomplete unusual gem behavior?

Felix Gallo — 2011-04-29T19:08:23

system: ubuntu linux 10.10
ruby: 1.9.2p180 manually compiled from source
vim: 7.3 manually compiled from source with +ruby option enabled
vim-ruby: latest installed, plus the 0.9 rubycomplete version from the
rubycomplete_updates branch.

$ vi x.rb
#!/usr/local/bin/ruby

x = String.new

x.^X^O  -> many correct methods found
^[:q!

$ sudo gem install redis
Successfully installed redis-2.2.0
1 gem installed
Installing ri documentation for redis-2.2.0...
Installing RDoc documentation for redis-2.2.0...

$ vi y.rb
#!/usr/local/bin/ruby
require 'redis'

x = Redis.new

x.^X^O  -> no methods found


Note that

p x.methods

does in fact show all of the methods one might hope from the Redis
class, so it's not a syntax error or no-method problem.

I started trying to debug rubycomplete but unfortunately left my
necronomicon in the tower of mad Alzhazred in the doomed isles and now
my house is infested with shoggoths; thanks Mark Guzman.

F.

Re: Abandoning Rubyforge

Czarek — 2011-01-07T20:36:58

Naaahhh... I think plain old CVS repo is better. Do RCS folks have any
objections?

Re: Abandoning Rubyforge

Mark Guzman — 2011-01-07T01:32:27

On Jan 4, 2011, at 12:41 PM, Tim Pope wrote:


+1

Re: omni-completion refresh

Mark Guzman — 2011-01-07T01:33:40

Jarmo,
  I've got a few changes in a branch called rubycomplete_updates that should handle the namespacing issue.
With regards to file reloading, I have some options I've been playing with involving moving a lot of the processing outside of the Vim process.
  --mg

On Dec 26, 2010, at 12:58 AM, Jarmo wrote:


_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Re: Abandoning Rubyforge

Tim Pope — 2011-01-04T17:41:35

+1 on moving to GitHub.

Re: Abandoning Rubyforge

Nikolai Weibull — 2011-01-04T16:54:03

No.  Go for it.

Abandoning Rubyforge

Doug Kearns — 2011-01-04T16:32:47

Does anyone have any objections to abandoning the RubyForge project, other than
redirects, for GitHub?  While it seems RubyForge now pretty much provides all
the things that motivated the partial move to GitHub in the first place it's,
likewise, apparent that GitHub now provides most of the features that were
required from RubyForge.

I'm indifferent to which site to use but it doesn't make sense using both
anymore.

Regards,
Doug

Re: omni-completion refresh

Jarmo — 2010-12-26T05:58:03

Hi!

Let me add some new findings about omni-completion when modules are in use.
Even the following snippet will show "pattern not found":
module A
  class B
    def c

    end
  end
end

a = A::B.new
a.#<C-x><C-o>

By setting < at >< at >debug to true in rubycomplete.vim i see these messages when
trying to complete for the first time:
input a.
variable
load_buffer_class(a) START
vartype A::B
load_buffer_class(A::B) START
load_buffer_class err: (eval):1:in `get_completions': uninitialized constant
VimRubyCompletion::A

For some reason module A is not eval'ed to create the constant. In
load_buffer_class the line:
  classdef = get_buffer_entity(name, 's:GetBufferRubyClass("%s")')

returns nil.

Is there any better way to debug the Ruby code within rubycomplete.vim than
with those dprint() statements? Any chance to use ruby-debug there somehow?

Should the code completion even work with today's Ruby? I can see from the
github history that there hasn't been much of an action lately with
rubycomplete.vim, but that could me

omni-completion refresh

Jarmo — 2010-12-25T17:44:35

Hi!

I've finally got Ruby code completion working sort of, but not as i'd really
like it. Currently it doesn't show the new methods added to some class if
code-completion has already used. For example, if i have files like these:

# a.rb
class A
  def my_method
  end
end

# b.rb
require "a"
a = A.new
a. # <C-x><C-o> will show my_method

Now if i open a.rb in a separate buffer and add some new method and save the
file then again in the b-buffer omni-completion doesn't show the newly added
method. I have to restart vim to make it show me the new method too. Is it
possible to make vim reload the omni-completion data?

In my vimrc i have currently these lines:
autocmd FileType ruby,eruby set omnifunc=rubycomplete#Complete
autocmd FileType ruby,eruby let g:rubycomplete_buffer_loading = 1
autocmd FileType ruby,eruby let g:rubycomplete_rails = 1
autocmd FileType ruby,eruby let g:rubycomplete_classes_in_global = 1

Also, when i use modules then the omni-completion doesn't seem to work at
all - even if i restart vim

Re: gvim editing ruby on Windows

Steve Horn — 2010-10-31T17:47:56

Had ruby 1.8.7 installed along with 1.9.1. Un-installed both and installed
1.9.2 and the error has gone away.
On Oct 31, 2010 11:40 AM, "Tim Pope" <vim-ruby-devel< at >tpope.info> wrote:
_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Re: gvim editing ruby on Windows

Tim Pope — 2010-10-31T14:46:48

Probably a Ruby 1.9 thing. What does `:ruby p RUBY_VERSION` output?

I know there's issues with the omnicomplete plugin. This is the first I've
heard of with just the ftplugin.

On Sat, Oct 30, 2010 at 10:19 PM, Steve Horn <steve< at >stevehorn.cc> wrote:

_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

gvim editing ruby on Windows

Steve Horn — 2010-10-31T02:19:38

When I try to edit an existing or create a new ruby file with gvim on
Windows I get the following error:
----
Error detected while processing C:\Program Files
(x86)\Vim\vim73\ftplugin\ruby.vim
line 76:
NoMethodError: undefined method 'synchronize' for #<Mutex:0x4019cc0>
line 93:
E121: Undefined variable: s:ruby_path
E15: Invalid expression: s:ruby_path

Any idea what I might be doing wrong?
_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Syntax highlighting of multiple here documents

Nikolai Weibull — 2010-10-26T16:49:53

p <<EOE, <<EOA
abc
EOE
def
EOA

isn’t highlighted correctly.  I can’t think of a way to good fix this
off the top of my head, but I guess one solution would be to allow for
here documents to match inside here documents, or match the last one
on a line (which I don’t think you can tell Vim to do).

Anyway, thought I’d mention it.  Also, I’m planning on upgrading the
indentation definition, but I’m not making any promises.
_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Fwd: Error by opening a ruby file with gvim 7.3.5

Gavin Sinclair — 2010-09-16T04:50:41

Hi folks,

Any thoughts on this problem that's been reported to me?

Gavin


---------- Forwarded message ----------
From: Cesar Romani <cesar.romani< at >gmail.com>
Date: Thu, Sep 16, 2010 at 12:42 AM
Subject: Error by opening a ruby file with gvim 7.3.5
To: gsinclair< at >gmail.com


Dear Gavin,
I'm using the latest release of vim, 7.3.5 compiled with MinGW on Win XP
with ruby support, version 1.9.1-p378-rc2
If I open whatever ruby file, I get the following error messages:

--------------------
Error detected while processing C:\Programmi\Vim\vim73\ftplugin\ruby.vim:
line   76:
NoMethodError: undefined method `synchronize' for #<Mutex:0xfeb520>
line   93:
E121: Undefined variable: s:ruby_path
E15: Invalid expression: s:ruby_path
--------------------

Best regards,

Cesar

--
Cesar

Re: Last Change headers

Nikolai Weibull — 2010-09-02T07:38:07


Well, I don’t know exactly, but I guess I decided on “Latest Revision”
about 150 runtime files ago and never looked back.  The date format
that I use is, hands down, the best there is.
_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Re: Last Change headers

Doug Kearns — 2010-09-01T21:28:51

I meant that yours is "Latest Revision: YYYY-MM-DD" vs the more 'standard'
"Last Change: YYY Month DD" or whatever it is.  I just thought there might be a
pertinent reason for that.


Thanks.

Doug

Re: Last Change headers

Nikolai Weibull — 2010-09-01T20:41:04


Unique?  Don’t all scripts have that stupid header?  Now that Vim is
finally under a decent, albeit inferior, SCM I really don’t see the
point of this anymore.

I use my own Vim plugin for this:

http://github.com/now/vim-templates

Note that it depends on:

http://github.com/now/vim-now-base
_______________________________________________
vim-ruby-devel mailing list
vim-ruby-devel< at >rubyforge.org
http://rubyforge.org/mailman/listinfo/vim-ruby-devel

Last Change headers

Doug Kearns — 2010-09-01T18:23:55

Does anyone have any thoughts on the best fix for this irritation?

Is it worth doing some sort of keyword expansion with smudge/clean filters?  I
think export-subst leaves $Format: ...$.  Perhaps a Vim specific release script
is the simplest approach since it only happens once every few years.

Nikolai, what are you using for your Vim scripts?  I notice you're going it
alone with a somewhat unique "Latest Revision" header.  How's that generated?

Doug

Re: Worth supporting 1.9 conditionally?

Doug Kearns — 2010-09-01T18:12:51

<snip>


Nothing interesting.  I just noticed some of the predefined constants/variables
were missing.  E.g. RUBY_ENGINE, __ENCODING__, __callee__ etc.


I just stuck it in rubyOperators for those fond of rainbows.  I don't really
think it needs highlighting otherwise.  "->" and "lambda" are roughly analogous
to "||" and "or".


Longer than myself then. :)

Doug